Creating a panel of normal from one normal sample

Question

[Jordan]

Hi,

I have just one normal sample and 3 tumors from the same patient. In this case is it ok to generate PoN using the following code?

gatk --java-options "-Xmx16g" Mutect2 \
    -R $ref \
    -I $BAMs/normal.bam \
    -tumor normal \
    --germline-resource somatic-hg38-af-only-gnomad.hg38.vcf \
    -O $out/normal_for_pon.vcf.gz

Can I use the normal_for_pon.vcf.gz for subsequent steps?

I couldn't find much information about using just one normal sample in the tutorials. I'm using gatk v4.1.1.1.

Thanks for the help!

Answer 1

[Kevin Blighe]

In this case, it is anything but a 'panel' of normals; so, calling it such would be misleading to any person reviewing your work. What you are essentially doing is an unpaired tumour-normal analysis, and the n for your normal group is just n=1, i.e., you are comparing each tumour to the same normal sample.

There are a few points about such an analysis that will prevent you from publishing this work in any reputable journal.

Kevin